The CEO’s Workout for Peak AI Performance

To continuously realize value from AI and GenAI for their companies, CEOs need to run both a sprint and a marathon. They must cultivate speed and strength to invest strategically, accelerate adoption, and secure world-class AI talent. At the same time, they must also build endurance to maintain momentum, remain agile, and master the art of imagining the possible with this dynamic technology.

The level of dedication required is no different than that of a top athlete. Just as a runner develops a training regimen to achieve world-class performance, CEOs need a “workout” routine to ensure their organizations achieve peak AI performance.

The need to get AI-fit is pressing. Only one in four executives surveyed by BCG say their companies are seeing significant returns from their AI and GenAI investments. And because AI fundamentally transforms the way people work, the longer laggards take to catch up, the more they risk falling permanently behind their competitors.

Put simply, this is not a “wait and see” moment. It’s a “go fast and accelerate” moment.

To get the most out of an AI workout, CEOs need to commit to using the technology every day and designate a personal AI fitness trainer to help them maintain discipline and rigor. They also need to start a two-way dialogue with their employees, both to help them understand the role AI plays in their company and to glean “boots on the ground” insights from those using it.

There is no end to the demands on a CEO’s time. But the benefits of an AI workout will make the effort well worth their while. Here’s how.

Benefit #1: Increasing Focus and Awareness

It’s been more than two years since ChapGPT propelled GenAI to the top of the CEO agenda. The technology has evolved exponentially since then, increasing the odds that leaders are basing their AI vision and strategy on outdated information.

Leaders need to know what AI and GenAI can do today if they’re going to make the most informed decisions about where to place their near-term portfolio bets while selecting longer-range investments that leverage frontier developments.

This is especially germane at this moment, because companies that wrestle with scaling AI and GenAI often spread their resources too thin by launching too many pilots. CEOs who are AI-fit can avoid this trap by more easily homing in on a handful of high-value initiatives that drive efficiencies and build long-term advantage. BCG research finds that leaders who take this approach scale 2.3 more AI products across their companies and achieve more than twice the return on their AI investments than their unfocused peers.

An AI workout also cultivates greater awareness for CEOs who want to stay abreast of what’s happening outside their company and the ecosystems in which they’re operating. The appearance of AI agents on CEO radars is a case in point: in the final quarter of 2024, mentions of AI agents on earnings calls increased more than 330% year over year, according to IoT Analytics data and BCG analysis.

That’s late in the game to be tuning into AI agents. CEOs could have started experimenting with off-the-shelf AI agents last May, when the first desktop agent was released.¹ 1 BCG was building AI agents in late 2023 and released an open-source toolkit a little over a year ago

Think what even a six-month head start would mean for a CEO looking to reshape critical functions or invent new products and services. They would have recognized the dramatic impact AI agents can have well before many of their competitors. They would have understood sooner that AI agents do multiple tasks autonomously—and therefore require more time for testing and learning than single AI assistants.

And because AI agents don’t just access data, but entire systems, CEOs who’ve been using the technology for months now would already appreciate how flaws in those systems could amplify the challenges of adopting and scaling agents.

Benefit #2: Assessing Risks and Building Resilience

When someone starts working with AI, they’re never quite sure what the output will be. Savvy prompt engineering is a new skill, and it takes a lot of iteration. The same holds true when designing a robust AI strategy. It takes hands-on experimentation and constant learning—two hallmarks of an AI workout—to accurately assess the risks of integrating the tech into an operation.

The workout can also help CEOs develop a people-centered blueprint to build, test, iterate, and scale the technology and ensure it remains flexible enough to pivot in a rapidly morphing AI landscape.

Most CEOs know that if someone in their organization doesn’t want something to happen, they’ll overplay the risks. By experimenting with GenAI themselves and staying on top of trends, CEOs can better distinguish risks that are acceptable from those that are not, rather than default to playing it safe (or charging ahead recklessly).

Consider again the example of AI agents. The more CEOs test them, the more fully they’ll appreciate the steep but worthwhile climb their company will face to achieve a smooth, trustworthy agentic workflow.

CEOs who roll up their sleeves and take a hands-on approach with AI will also gain a much deeper understanding of how AI challenges and changes the way people work—helping them develop strong, people-centered strategies. That’s the lion’s share of BCG’s 10-20-70 formula for AI success: 10% of resources should be devoted to building algorithms, 20% to high quality data and tech, and 70% to transforming the way people and processes operate.

Experimenting with different models can also help CEOs cultivate a healthier degree of skepticism before locking into a trajectory they can’t easily adjust. Depending on where a company is in its AI journey, 40,000 licenses from an AI provider might make sense—or not. Given the velocity of change with AI and GenAI, the model that is best suited to a company’s needs today could be displaced by a better option tomorrow. A more flexible strategy also builds resilience against regulatory and geopolitical shifts that can impact access to models, data, and highly coveted talent.

Benefit #3: Demystifying AI and Securing Greater Buy In

CEOs who use AI and GenAI themselves are not only more likely to develop smarter strategies and boost adoption—they can also help external and internal stakeholders better understand the value of these investments.

By sharing their personal experiences of working with AI, CEOs can help demystify the tech for frontline employees, allowing them to embrace its transformative potential with greater optimism. In doing so, CEOs can also help embed a responsible AI culture. A study by Slack, for instance, found that nearly half of all desk workers would feel uncomfortable disclosing to their managers that they use AI for simple tasks. When workers know their managers and senior leaders—especially the CEO—are using AI every day, it destigmatizes the technology, encouraging employees who are using it clandestinely to come out of the shadows.

Finally, an AI workout can help CEOs attract top AI talent. BCG research has found that the most sought-after GenAI experts place a high premium on learning opportunities and interactions with executives on high-priority projects. If a company is led by someone who doesn’t understand AI, lacks a credible strategy, or is biding their time to see how their competitors’ AI investments fare, it’s unlikely to create the kind of environment where the best AI talent wants to work.

The CEO’s AI Workout: Four Essential Steps

Like any fitness routine, an AI workout should be built around a CEO’s individual goals and needs. But these four essential steps should be a part of every leader’s training regimen.

Commit to working with AI and/or GenAI every day. An effective AI workout starts with using the technology every day, no exceptions. It’s no different than a computer or smartphone. CEOs can even think of AI as a digital chief of staff that they confer with regularly throughout the day.

Designate a personal AI fitness trainer. CEOs don’t have time to test every new model and digest the ever-growing volume of information on AI and its surrounding ecosystems. A personal AI fitness trainer can help leaders focus on what matters most. The trainer can identify the most promising tools, learn how to use them, and help the CEO get caught up on them quickly. The trainer can also curate information to make sure only the most relevant items land on the CEO’s reading list—and can act as an accountability partner to ensure the CEO is adhering to the workout.

The personal AI fitness trainer should come from inside the CEO’s organization. They don’t necessarily need a degree in computer science or engineering—what matters is that they are a heavy AI user, someone who loves playing with all the new models and is passionately seeking out all the latest trends. They’re the kind of person who waits with bated breath when a new model is released on a Friday and has already spent twenty hours testing it come Monday.

Know the fundamentals—and keep learning. It would be tough to imagine an airline pilot climbing into the cockpit without having gained some understanding of how an aircraft works. Or a ship’s captain heading out to sea without understanding the basics of how a well-designed boat is engineered. The same applies to leading an AI transformation. CEOs don’t need to have a deep, granular understanding of the technology, but they should know the fundamentals—including basic terminology, how AI systems are created, and how they learn and become more intelligent. (See the sidebar.) From there, leaders can start working with the technology and developing their understanding of what it can really do.

For example, only when a CEO has explored how an AI agent can help with simple tasks—and then more complex tasks—can they imagine how the tech could work across an entire function. This type of approach builds confidence at every milestone, helping CEOs more fully envision how AI can transform their company.

Talk about “the why” and listen to employees. CEOs need to help their employees overcome their doubts about working with AI and GenAI by connecting the technology to the broader strategy and purpose of the company. Put simply, it is the CEO’s job to explain in clear terms “the greater why” of AI and its benefits to the organization.

When employees do start working with the technology, CEOs can ask them directly about their experiences with it. This will help them identify lessons and use cases that can help other parts of the organization and gain even more insights for connecting AI to the broader company purpose and employee value proposition. As the technology scales, CEOs can also hold town halls to get direct feedback and encourage employees to share experiences.

A Final Word on Getting and Staying in Shape

It may be early days in the AI revolution, but it’s high time CEOs get “transformation fit” by using the technology every day and staying current on important developments. It’s not something they need to do alone. With the help of a personal AI fitness trainer, they can make more informed decisions around their AI and GenAI investments and initiatives, give their employees greater encouragement to embrace the technology, and manage shareholder expectations more effectively.

For leaders of companies that are lagging on AI transformations, the workout can help them accelerate their efforts. And for CEOs who are already ahead of the curve, they can achieve an even more commanding lead.